Machine Learning Based Diabetic Prediction Using Random Forest

In recent years, the global impact on diabetics has increased, which is a significant issue. In this case, the patient is obliged to visit a diagnostic centre persistently to get their reports and after consultation investing time and currency on it will be inconvenient. Because of these reasons, outcomes may be severe if unnoticed. An increase in machine learning approaches solves this crucial disadvantage. The objective of this study is to create a method that helps to achieve an early prediction of diabetics with higher precision using random forest algorithm. The degree of precision is higher than other algorithms, with random forest we achieved an accuracy of 85.6% and found to be better algorithm for diabetic prediction comparing with other algorithms such as logistic regression, Naive Bayes, Gradient boosting classifier, KNN and SVM. Random forest yields effective outcomes for predicting diabetics and the result showed that the predictive method can predict the diabetics.